Intelligent MICC cabinet control system with leakage monitoring function
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ANHUI RUICHENG ENG TECH CO LTD
- Filing Date
- 2026-04-15
- Publication Date
- 2026-07-10
Smart Images
Figure CN122363155A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of MICC cabinet control system technology, specifically to an intelligent MICC cabinet control system with leakage monitoring function. Background Technology
[0002] With the rapid development of industrial automation and precision manufacturing technologies, modular integrated control cabinets are increasingly widely used in semiconductor manufacturing, data center cooling, chemical experiments, medical equipment and other fields. The complex liquid circuit system deployed inside makes leakage a key hidden danger affecting the safe operation of the equipment.
[0003] Currently, leakage monitoring solutions for MICC cabinets primarily employ a single sensor network, using a uniform monitoring strategy across all monitoring areas. This fails to differentiate based on area importance, equipment value, and leakage risk level, resulting in insufficient guarantees for rapid response in critical, leak-prone areas. Furthermore, excessive monitoring of non-critical areas wastes system resources. In addition, existing solutions lack dynamic adaptability. MICC cabinets have diverse application scenarios, and equipment deployment changes or business adjustments may occur during use. Traditional solutions fix the monitoring strategy after installation, failing to dynamically adjust judgment criteria and handling strategies according to changing scenarios, severely impacting the accuracy and effectiveness of leakage monitoring.
[0004] Therefore, an intelligent MICC cabinet control system with leakage detection function is needed to improve the above problems. Summary of the Invention
[0005] The purpose of this invention is to provide an intelligent MICC cabinet control system with leakage monitoring function to solve the problems mentioned in the background art.
[0006] To achieve the above objectives, the present invention provides the following technical solution: A smart MICC cabinet control system with leakage detection function includes: A high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network are provided. The high-sensitivity leakage monitoring network is used for real-time monitoring and rapid handling of core leakage-prone areas inside the cabinet, while the low-sensitivity leakage monitoring network is used for routine leakage inspection of non-core areas inside the cabinet. A leakage detection model corresponding to the application scenario of the MICC cabinet. The leakage detection model is constructed by a wide neural network. Its input corresponds one-to-one with the system perception unit. The application scenario indicates the usage type of the MICC cabinet and the business type of the equipment or pipelines deployed in the cabinet. The high-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response conditions for leakage based on the leakage judgment model. If so, the handling strategy is executed; otherwise, it switches to the low-sensitivity leakage monitoring network. The low-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response leakage conditions based on the leakage judgment model. If so, it is added to the high-sensitivity leakage monitoring network and emergency response is performed; otherwise, routine inspection is performed.
[0007] As a preferred embodiment of the invention, it includes: A high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network are provided. The high-sensitivity leakage monitoring network is used for real-time monitoring and rapid handling of core leakage-prone areas inside the cabinet, while the low-sensitivity leakage monitoring network is used for routine leakage inspection of non-core areas inside the cabinet. A leakage detection model corresponding to the application scenario of the MICC cabinet. The leakage detection model is constructed by a wide neural network. Its input corresponds one-to-one with the system perception unit. The application scenario indicates the usage type of the MICC cabinet and the business type of the equipment or pipelines deployed in the cabinet. The high-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response conditions for leakage based on the leakage judgment model. If so, the handling strategy is executed; otherwise, it switches to the low-sensitivity leakage monitoring network. The low-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response leakage conditions based on the leakage judgment model. If so, it is added to the high-sensitivity leakage monitoring network and emergency response is performed; otherwise, routine inspection is performed.
[0008] As a preferred embodiment of the invention, the main sensing unit is a high-precision leakage monitoring component, including a contact leakage sensing rope, a non-contact liquid level sensor and a high-precision temperature and humidity sensor, which are deployed in core leakage-prone areas such as pipe joints, liquid valve groups and the bottom of the equipment inside the cabinet, with a sampling frequency of 5Hz and a leakage signal sampling accuracy of 0.05M. The sensing unit is a conventional leakage monitoring component, including a patch-type leakage sensor and a common temperature and humidity sensor, deployed in a non-core area inside the cabinet, with a sampling frequency of 1Hz and a leakage signal sampling accuracy of 0.1M.
[0009] As a preferred embodiment of the invention, the step of determining whether the monitoring area data meets the high response condition for leakage specifically involves: Input the signal amplitude parameters, signal duration parameters, ambient temperature and humidity parameters, and area equipment correlation parameters corresponding to the leakage sensing data into the leakage judgment model to obtain the leakage risk level value. The high response condition for leakage is that the leakage risk level value is greater than or equal to a preset threshold.
[0010] As a preferred embodiment of the invention, the leakage judgment model configured for the application scenario of the MICC cabinet is specifically as follows: Set description parameters that match the current application scenario to identify the monitoring area type of the MICC cabinet, the service type of the liquid circuit equipment in the cabinet, and the level of leakage impact. The input to the width neural network includes leakage detection parameters and scene configuration parameters, and the output is the leakage risk level value; A training sample containing parameters corresponding to different leakage risk levels and actual handling results was constructed, and a leakage judgment model was obtained by training a wide neural network. When the application scenario of the MICC cabinet or the deployment of equipment inside the cabinet changes, the leakage detection model should be reconfigured.
[0011] As a preferred embodiment of the invention, the system sets the monitoring range of the high- and low-sensitivity leakage monitoring networks, specifically as follows: The communication latency between the sensing unit and the edge computing core node is calculated, and the leakage impact weight of the sensing unit deployment area is obtained. The leakage impact weight is set according to the value of the equipment in the area, the importance of the liquid path, and the risk of leakage spread. If the communication delay is preset to a delay threshold and the leakage impact weight is preset to a weight threshold, then the sensing unit is assigned to the leakage high-sensitivity monitoring network; otherwise, it is assigned to the leakage low-sensitivity monitoring network. The system is configured with a monitoring range update cycle. After the cycle is met, the parameters are recalculated and the monitoring range is updated.
[0012] As a preferred embodiment of the invention, the main execution control unit includes a high-speed solenoid valve, an emergency solenoid lock, a high-decibel audible and visual alarm, and a fire-fighting linkage dry contact, with an action response delay of 1 second, which is suitable for rapid emergency response under high-response conditions of leakage. The actuator control unit includes a conventional solenoid valve and a common audible and visual alarm, with a 3-second response delay, suitable for conventional alarms and handling under low response conditions due to leakage.
[0013] As a preferred embodiment of the invention, the system further includes a computer vision perception sub-network, which includes high-definition cameras inside and outside the cabinet, an image analysis unit, and data connections with a high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network. The YOLOv8n target detection model and Mask R-CNN instance segmentation model were used to identify the leakage spread range and abnormal equipment status inside the cabinet. The visual analysis parameters were then input into the leakage judgment model as the basis for calculating the leakage risk level.
[0014] As a preferred embodiment of the invention, the system further includes a computer vision perception sub-network, which includes high-definition cameras inside and outside the cabinet, an image analysis unit, and data connections with a high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network. The YOLOv8n target detection model and Mask R-CNN instance segmentation model were used to identify the leakage spread range and abnormal equipment status inside the cabinet. The visual analysis parameters were then input into the leakage judgment model as the basis for calculating the leakage risk level.
[0015] As a preferred embodiment of the invention, the system further includes a cloud management unit and a status feedback closed-loop unit. The cloud management unit communicates with the edge computing core node of the leakage high-sensitivity monitoring network and is used for monitoring data storage, iterative optimization of the leakage judgment model, remote monitoring and alarm push, and synchronizes the optimized model to all edge computing nodes. The status feedback closed-loop unit is used to collect the action status of the execution control unit and the subsequent monitoring data of the sensing unit in real time and feed them back to the leakage judgment model. The model recalculates the leakage risk level value based on the feedback data. If it is lower than the threshold, a reset command is sent. If it is continuously higher than the threshold, an upgrade disposal command is sent.
[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves differentiated monitoring and handling of core leak-prone areas and non-core areas within the cabinet by constructing a high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network. The system uses a dynamic monitoring range division mechanism to automatically determine the network to which the sensing unit belongs based on the communication latency of the sensing unit and the leakage impact weight, and sets an update cycle for dynamic adjustment, enabling the system to achieve optimal allocation of monitoring resources while ensuring the safety of the core area. This invention introduces a leakage judgment model based on a wide neural network, enabling intelligent and multi-dimensional comprehensive judgment of leakage risk. The system integrates a computer vision perception sub-network, which identifies the leakage spread range and abnormal equipment status through target detection and instance segmentation models. Visual analysis parameters are fused into the judgment model to further improve decision-making accuracy. The setting of a cloud management unit and a status feedback closed-loop unit enables continuous iterative optimization of the model and closed-loop verification of the handling effect, allowing the system to adapt to different application scenarios and equipment changes, thereby improving the accuracy and effectiveness of leakage monitoring. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the overall system principle of the present invention; Figure 2 This is a block diagram illustrating the internal workings of the core module of this invention. Figure 3 This is a block diagram illustrating the linkage principle of the present invention. Detailed Implementation
[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the invention. Obviously, the described embodiments are only some embodiments of the invention, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without creative effort are within the scope of protection of the present invention.
[0019] To facilitate understanding of the invention, a more complete description of the invention will be given below with reference to relevant embodiments. Several embodiments of the invention are provided. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
[0020] Please see Figure 1-3 The present invention provides a technical solution: For an example, please refer to... Figure 1 , 2 3. An intelligent MICC cabinet control system with leakage detection function, including a high-sensitivity leakage detection network and a low-sensitivity leakage detection network. The high-sensitivity leakage detection network is used for real-time monitoring and rapid handling of core leakage-prone areas inside the cabinet, and the low-sensitivity leakage detection network is used for routine leakage inspection of non-core areas inside the cabinet. A leakage detection model corresponding to the application scenario of the MICC cabinet. The leakage detection model is constructed by a wide neural network. Its input corresponds one-to-one with the system perception unit. The application scenario indicates the usage type of the MICC cabinet and the business type of the equipment or pipelines deployed in the cabinet. The high-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response conditions for leakage based on the leakage judgment model. If so, it executes the handling strategy; otherwise, it switches to the low-sensitivity leakage monitoring network. The low-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response leakage conditions based on the leakage judgment model. If so, it is added to the high-sensitivity leakage monitoring network and emergency response is performed; otherwise, routine inspection is performed.
[0021] In this embodiment of the invention, the wide neural network can adopt a three-layer structure, namely a first hidden layer, a second hidden layer, and a third hidden layer. The activation function of the first hidden layer is the Mish function, and batch normalized dropout technology is used to randomly delete neurons in the first hidden layer. The activation function of the second hidden layer is the Swish function, and residual connection technology is used. Softmax and sigmoid functions are set in the second hidden layer to achieve attention gating. The activation function of the third hidden layer is the LeakyReLU(0.2) function, and grouped convolution is used to process spatial features. The input layer dimension corresponds one-to-one with all perceptual units in the system, and the output layer is the leakage risk level value.
[0022] The training sample set includes parameter combinations of the sensor under different leakage conditions and corresponding historical handling results. For example, when the resistance value of the contact leakage sensing rope drops to 0.5MΩ, the ambient humidity rises to 80%, and the equipment value in the area is high, the corresponding leakage risk level is 0.9, and the handling strategy is to immediately close the liquid supply valve and trigger an audible and visual alarm.
[0023] During system operation, the high-sensitivity leakage monitoring network and the low-sensitivity leakage monitoring network work in parallel and switch dynamically according to the judgment model. The high-sensitivity leakage monitoring network includes a main sensing unit, an edge computing core node, and a main execution control unit. The main sensing unit and the edge computing core node are directly connected via an industrial high-speed communication protocol, and the edge computing core node is used to control the main execution control unit.
[0024] The main sensing unit is a high-precision leak detection component, including a contact-type leak sensing rope, a non-contact liquid level sensor, and a high-precision temperature and humidity sensor. It is deployed in key leak-prone areas such as pipe joints, liquid valve assemblies, and the bottom of the equipment within the cabinet. The sampling frequency is 5Hz, and the leak signal sampling accuracy is 0.05M. The main execution control unit includes a high-speed solenoid valve, an emergency solenoid lock, a high-decibel audible and visual alarm, and a fire-fighting linkage dry contact. The action response delay is 1 second, adaptable to rapid emergency response under high-response conditions of leaks.
[0025] Please refer to Figure 2 The low-sensitivity leakage monitoring network comprises sensing units, edge computing slave nodes, and slave execution control units. Multiple sensing units interact with the edge computing core node of the high-sensitivity leakage monitoring network via edge computing slave nodes. The slave execution control units are centrally scheduled by the edge computing core node. The sensing units are conventional leakage monitoring components, including patch-type leakage sensors and ordinary temperature and humidity sensors, deployed in non-core areas within the cabinet, with a sampling frequency of 1Hz and a leakage signal sampling accuracy of 0.1M. The slave execution control units include conventional solenoid valves and ordinary audible and visual alarms, with a 3s response delay, suitable for routine alarms and handling under low-response leakage conditions.
[0026] The operation process of the highly sensitive leakage monitoring network is as follows: the main sensing unit collects data from the core leakage-prone area in real time and connects directly to the edge computing core node through the Profinet industrial Ethernet protocol. The edge computing core node inputs the received signal amplitude parameters, signal duration parameters, environmental temperature and humidity parameters, and regional equipment correlation parameters into the leakage judgment model to obtain the leakage risk level value.
[0027] If the leakage risk level value is greater than or equal to the preset threshold, it is determined that the leakage high response condition is met. The edge computing core node directly drives the main execution control unit to perform rapid emergency response, while the sensing unit continues to remain in the high-sensitivity monitoring network. If the leakage risk level value is less than the preset threshold, it is determined that the leakage high response condition is not met, the sensing unit exits the high-sensitivity monitoring network, its monitoring task is downgraded, and it switches to the low-sensitivity leakage monitoring network.
[0028] The operation process of the low-sensitivity leakage monitoring network is as follows: data from non-core areas is collected from sensing units, aggregated from edge computing nodes, and interacts with the edge computing core node via the Modbus-RTU protocol. The edge computing core node inputs the received parameters into the leakage judgment model to obtain the leakage risk level value. If the leakage risk level value is greater than or equal to a preset threshold, it is determined to meet the high-response leakage condition. The edge computing core node immediately adds the sensing unit to the high-sensitivity monitoring network, including dynamically switching the unit's communication protocol to the Profinet industrial Ethernet protocol, increasing its sampling frequency to 5Hz, and including it in the fast response scheduling queue of the main execution control unit, and then performing emergency response. If the leakage risk level value is less than the preset threshold, it is determined that the high response condition for leakage is not met. The sensing unit continues to remain in the low-sensitivity monitoring network. The edge computing core node only retains the data, and the execution control unit performs routine alarms and handling.
[0029] In a preferred embodiment of the present invention, the system further defines the monitoring range of high- and low-sensitivity leakage monitoring networks. Specifically, the system first calculates the communication latency between the sensing unit and the edge computing core node, and obtains the leakage impact weight of the sensing unit deployment area. This leakage impact weight is set based on the value of the equipment in the area, the importance of the liquid path, and the risk of leakage spread.
[0030] If the communication delay is less than or equal to the preset delay threshold, and the leakage impact weight is greater than or equal to the preset weight threshold, then the sensing unit is assigned to the high-sensitivity leakage monitoring network; otherwise, it is assigned to the low-sensitivity leakage monitoring network.
[0031] Meanwhile, the system is configured with a monitoring range update cycle. After the update cycle is met, the system recalculates the above parameters and updates the monitoring range to adapt to changes in the network environment or changes in the equipment inside the cabinet.
[0032] In another preferred embodiment of the invention, the system also integrates a computer vision perception sub-network. This sub-network includes high-definition cameras and image analysis units installed inside and outside the cabinet, and is connected to a high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network for data transmission.
[0033] In practical applications, when a leak occurs, a sensor in the highly sensitive monitoring network triggers an alarm, and the system immediately activates the in-cabinet camera to focus on the affected area. The image analysis unit uses the YOLOv8n object detection model to identify the liquid accumulation area in the image in real time, and uses the Mask R-CNN instance segmentation model to accurately segment the liquid diffusion contour, calculating the leak area and diffusion rate. These visual analysis parameters, such as the proportion of the leak diffusion area to the chassis area and the leak diffusion rate, are input as new feature parameters into the leak detection model.
[0034] The model combines existing sensor data with visual data to comprehensively assess a more accurate leakage risk level. For example, even if the sensors detect a small amount of liquid, but visual analysis indicates that the leakage is rapidly spreading towards high-value circuit boards, the model's risk level will significantly increase, triggering a higher-level response order.
[0035] In another preferred embodiment of the present invention, the system is configured with dual communication protocols: the high-sensitivity leakage monitoring network adopts the Profinet industrial Ethernet protocol to achieve low-latency and high-reliability data transmission, and the low-sensitivity leakage monitoring network adopts the Modbus-RTU protocol to balance communication efficiency and deployment cost.
[0036] When a sensing unit in the low-sensitivity leakage monitoring network detects data that meets the high-response leakage conditions, the edge computing core node automatically switches the communication protocol of the sensing unit to the Profinet Industrial Ethernet protocol and incorporates it into the high-sensitivity leakage monitoring network.
[0037] For example, a patch sensor that originally used the Modbus-RTU protocol has its communication protocol dynamically switched to the Profinet protocol after detecting a leakage signal and the risk level value calculated by the judgment model exceeds the threshold. The sampling frequency is increased from 1Hz to 5Hz, and it is added to the fast response queue of a high-sensitivity monitoring network.
[0038] In another preferred embodiment of the invention, the system further includes a cloud management unit and a status feedback closed-loop unit. The cloud management unit communicates with the edge computing core node of the leakage high-sensitivity monitoring network, and is used for monitoring data storage, iterative optimization of the leakage judgment model, remote monitoring and alarm push, and synchronizes the optimized model to all edge computing nodes.
[0039] For example, the cloud management unit uses a large amount of accumulated real leakage event data to periodically iterate and optimize the leakage judgment model. The optimized model parameters are synchronized back to all edge computing nodes via OTA, enabling the model to continuously evolve.
[0040] The status feedback closed-loop unit is used to collect the action status of the execution control unit and the subsequent monitoring data of the sensing unit in real time, and feed it back to the leakage judgment model. The model recalculates the leakage risk level value based on the feedback data. If it is lower than the threshold, a reset command is sent; if it is consistently higher than the threshold, an escalation command is sent. After the main execution control unit performs the action of closing the high-speed solenoid valve, the status feedback closed-loop unit will collect the valve core position status signal of the solenoid valve in real time and continue to monitor the subsequent data of the upstream sensing unit. This feedback data is sent back to the edge computing core node in real time, and the leakage judgment model recalculates the risk level value.
[0041] If the calculated new risk level value is lower than the reset threshold, the system will send a reset command to clear the alarm and attempt to restore some functions. If the risk level remains above the threshold, the system will send an escalation order, such as activating backup plans, notifying more personnel, or implementing more thorough isolation procedures.
[0042] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A smart MICC cabinet control system with leakage detection function, characterized in that, include: A high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network are used for real-time monitoring and rapid handling of core leakage-prone areas inside the cabinet, and the low-sensitivity leakage monitoring network is used for routine leakage inspection of non-core areas inside the cabinet. A leakage detection model corresponding to the application scenario of the MICC cabinet. The leakage detection model is constructed by a wide neural network. Its input corresponds one-to-one with the system perception unit. The application scenario indicates the usage type of the MICC cabinet and the business type of the equipment or pipelines deployed in the cabinet. The high-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response conditions for leakage based on the leakage judgment model. If so, it executes the handling strategy; otherwise, it switches to the low-sensitivity leakage monitoring network. The low-sensitivity leakage monitoring network determines whether the data in the monitoring area meets the high-response leakage conditions based on the leakage judgment model. If so, it is added to the high-sensitivity leakage monitoring network and emergency response is performed; otherwise, routine inspection is performed.
2. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The highly sensitive leakage monitoring network includes: The system comprises a main sensing unit, an edge computing core node, and a main execution control unit. The main sensing unit and the edge computing core node are directly connected via an industrial high-speed communication protocol. The edge computing core node is used to control the main execution control unit. The low-sensitivity leakage monitoring network includes a sensing unit, an edge computing slave node, and a slave execution control unit. Multiple sensing units interact with the edge computing core node of the high-sensitivity leakage monitoring network through the edge computing slave node. The slave execution control unit is uniformly scheduled by the edge computing core node.
3. The intelligent MICC cabinet control system with leakage monitoring function according to claim 2, characterized in that: The main sensing unit is a high-precision leakage monitoring component, including a contact leakage sensing rope, a non-contact liquid level sensor, and a high-precision temperature and humidity sensor. It is deployed in core leakage-prone areas such as pipe joints, liquid valve groups, and the bottom of the equipment inside the cabinet. The sampling frequency is 5Hz, and the leakage signal sampling accuracy is 0.05M. The sensing unit is a conventional leakage monitoring component, including a patch-type leakage sensor and a common temperature and humidity sensor, deployed in a non-core area inside the cabinet, with a sampling frequency of 1Hz and a leakage signal sampling accuracy of 0.1M.
4. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that... The determination of whether the monitoring area data meets the high response condition for leakage is specifically as follows: Input the signal amplitude parameters, signal duration parameters, ambient temperature and humidity parameters, and area equipment correlation parameters corresponding to the leakage sensing data into the leakage judgment model to obtain the leakage risk level value. The high response condition for leakage is that the leakage risk level value is greater than or equal to a preset threshold.
5. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The leakage detection model corresponding to the MICC cabinet application scenario is as follows: Set description parameters that match the current application scenario to identify the monitoring area type of the MICC cabinet, the service type of the liquid circuit equipment in the cabinet, and the level of leakage impact. The input to the width neural network includes leakage detection parameters and scene configuration parameters, and the output is the leakage risk level value; A training sample containing parameters corresponding to different leakage risk levels and actual handling results was constructed, and a leakage judgment model was obtained by training a wide neural network. When the application scenario of the MICC cabinet or the deployment of equipment inside the cabinet changes, the leakage detection model should be reconfigured.
6. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The system sets the monitoring range of the high-sensitivity and low-sensitivity leakage monitoring networks, specifically as follows: The communication latency between the sensing unit and the edge computing core node is calculated, and the leakage impact weight of the sensing unit deployment area is obtained. The leakage impact weight is set according to the value of the equipment in the area, the importance of the liquid path, and the risk of leakage spread. If the communication delay is preset to a delay threshold and the leakage impact weight is preset to a weight threshold, then the sensing unit is assigned to the leakage high-sensitivity monitoring network; otherwise, it is assigned to the leakage low-sensitivity monitoring network. The system is configured with a monitoring range update cycle. After the cycle is met, the parameters are recalculated and the monitoring range is updated.
7. The intelligent MICC cabinet control system with leakage monitoring function according to claim 2, characterized in that: The main execution control unit includes a high-speed solenoid valve, an emergency solenoid lock, a high-decibel audible and visual alarm, and a fire-fighting linkage dry contact. The action response delay is 1 second, which is suitable for rapid emergency response under high-response conditions of leakage. The actuator control unit includes a conventional solenoid valve and a common audible and visual alarm, with a 3-second response delay, suitable for conventional alarms and handling under low response conditions due to leakage.
8. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The system also includes a computer vision perception sub-network, which includes high-definition cameras inside and outside the cabinet, an image analysis unit, and data connections with a high-sensitivity leakage monitoring network and a low-sensitivity leakage monitoring network. The YOLOv8n target detection model and Mask R-CNN instance segmentation model were used to identify the leakage spread range and abnormal equipment status inside the cabinet. The visual analysis parameters were then input into the leakage judgment model as the basis for calculating the leakage risk level.
9. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The system is configured with dual communication protocols: the high-sensitivity leakage monitoring network adopts the Profinet industrial Ethernet protocol, and the low-sensitivity leakage monitoring network adopts the Modbus-RTU protocol. When a sensing unit in the low-sensitivity leakage monitoring network detects data that meets the high-response leakage conditions, the edge computing core node switches the communication protocol of the sensing unit to the Profinet Industrial Ethernet protocol and incorporates it into the high-sensitivity leakage monitoring network.
10. The intelligent MICC cabinet control system with leakage monitoring function according to claim 1, characterized in that: The system also includes a cloud management unit and a status feedback closed-loop unit. The cloud management unit communicates with the edge computing core node of the leakage high-sensitivity monitoring network and is used for monitoring data storage, leakage judgment model iterative optimization, remote monitoring and alarm push, and synchronizes the optimized model to all edge computing nodes. The status feedback closed-loop unit is used to collect the action status of the execution control unit and the subsequent monitoring data of the sensing unit in real time and feed them back to the leakage judgment model. The model recalculates the leakage risk level value based on the feedback data. If it is lower than the threshold, a reset command is sent. If it is continuously higher than the threshold, an upgrade disposal command is sent.